Sur d’autres sites (7778)

• avfilter/vf_lut3d : add x86-optimized tetrahedral interpolation

  6 octobre 2021, par Mark Reid

  avfilter/vf_lut3d : add x86-optimized tetrahedral interpolation
  I spotted an interesting pattern that I didn't see before that leads to the implementation being faster.
  
  The bit shifting table I was using before is no longer needed, and was able to remove quite a few lines. 
  
  I also add use of FMA on the AVX2 version.
  f32 1920x1080 1 thread with prelut
  
  c impl
  
  1434012700 UNITS in lut3d->interp,   1 runs,   0 skips
  
  1434035335 UNITS in lut3d->interp,   2 runs,   0 skips
  
  1423615347 UNITS in lut3d->interp,   4 runs,   0 skips
  
  1426268863 UNITS in lut3d->interp,   8 runs,   0 skips
  sse2
  
  905484420 UNITS in lut3d->interp,   1 runs,   0 skips
  
  905659010 UNITS in lut3d->interp,   2 runs,   0 skips
  
  915167140 UNITS in lut3d->interp,   4 runs,   0 skips
  
  915834222 UNITS in lut3d->interp,   8 runs,   0 skips
  avx
  
  574794860 UNITS in lut3d->interp,   1 runs,   0 skips
  
  581035090 UNITS in lut3d->interp,   2 runs,   0 skips
  
  584116720 UNITS in lut3d->interp,   4 runs,   0 skips
  
  581460290 UNITS in lut3d->interp,   8 runs,   0 skips
  avx2
  
  301698880 UNITS in lut3d->interp,   1 runs,   0 skips
  
  301982880 UNITS in lut3d->interp,   2 runs,   0 skips
  
  306962430 UNITS in lut3d->interp,   4 runs,   0 skips
  
  305472025 UNITS in lut3d->interp,   8 runs,   0 skips
  gbrap16 1920x1080 1 thread with prelut
  
  c impl
  
  1480894840 UNITS in lut3d->interp,   1 runs,   0 skips
  
  1502922990 UNITS in lut3d->interp,   2 runs,   0 skips
  
  1496114307 UNITS in lut3d->interp,   4 runs,   0 skips
  
  1492554551 UNITS in lut3d->interp,   8 runs,   0 skips
  sse2
  
  980777180 UNITS in lut3d->interp,   1 runs,   0 skips
  
  986121520 UNITS in lut3d->interp,   2 runs,   0 skips
  
  986489840 UNITS in lut3d->interp,   4 runs,   0 skips
  
  998832248 UNITS in lut3d->interp,   8 runs,   0 skips
  avx
  
  622212360 UNITS in lut3d->interp,   1 runs,   0 skips
  
  622981160 UNITS in lut3d->interp,   2 runs,   0 skips
  
  645396315 UNITS in lut3d->interp,   4 runs,   0 skips
  
  641057075 UNITS in lut3d->interp,   8 runs,   0 skips
  avx2
  
  321336400 UNITS in lut3d->interp,   1 runs,   0 skips
  
  321268920 UNITS in lut3d->interp,   2 runs,   0 skips
  
  323459895 UNITS in lut3d->interp,   4 runs,   0 skips
  
  324949967 UNITS in lut3d->interp,   8 runs,   0 skips
  

  • [DH] libavfilter/lut3d.h
  • [DH] libavfilter/vf_lut3d.c
  • [DH] libavfilter/x86/Makefile
  • [DH] libavfilter/x86/vf_lut3d.asm
  • [DH] libavfilter/x86/vf_lut3d_init.c

• Decoding AAC audio with ffmpeg

  21 janvier 2021, par Michael Brown

  I'm trying to decode an AAC audio stream in an ADTS container, which is streamed from an external hardware H264 encoder.

  


  I've parsed out the ADTS and it tells me I've got a 2 channel, 44100 AAC Main profile frame. I setup the extra data bytes for the ffmpeg decoder and decode the frame successfully ? as follows :

  


  (pseudo c++ code)

  


  setup the decoder :

  


  avcodec_find_decoder(codec_id);
avcodec_alloc_context3(context->codec);
avcodec_open2(context->av_codec_context, context->codec, nullptr);
av_init_packet(&context->av_raw_packet);


  


  setup the extra data bytes :

  


  // AOT_MAIN, 44.1kHz, Stereo
// 00001010 00010000
// extradata = 0x0A, 0X10
memcpy(context->av_codec_context->extradata, extradata, extradataLength);
avcodec_open2(context->av_codec_context, context->codec, nullptr);


  


  then decode the frame :

  


  // decode frame
const int len = avcodec_decode_audio4(context->av_codec_context, context->frame, &got_frame, &context->av_raw_packet);
*sampleRate = context->av_codec_context->sample_rate;
*sampleFormat = context->av_codec_context->sample_format;
*bitsPerSample = av_get_bytes_per_sample(context->av_codec_context->sample_fmt) * 8;
*channels = context->av_codec_context->channels;
*channelLayout = context->av_codec_context->channelLayout;
// get frame
*outDataSize = av_samples_get_buffer_size(nullptr, context->av_codec_context->channels, context->frame->nb_samples, context->av_codec_context->sample_fmt, 1);


  


  The decoded frame :

  


  // array of 8192 bytes, context info is as expected:
context->av_codec_context->channels = 2
context->av_codec_context->channelLayout = 3 (AV_CH_LAYOUT_STEREO)
context->frame->sample_fmt = 8 (AV_SAMPLE_FMT_FLTP) // float, planar
context->frame->sample_rate = 44100


  


  Now as I understand it each frame in the raw format for 32 bit will be 4 bytes per sample, and each channel will be interleaved (so every 4th byte is the alternating channel). That leaves me with 1024 samples for each channel (8192 / 32 bits / 2 channels).

  


  I've tried exporting multiple frames of this data to a file, and importing as a raw file (32-bit float, 2 channel 44100Hz, little endian) in Audacity to sanity check. Instead of music, all I get is noise and the detected length of the audio is way longer than I would have expected (5 seconds dumped to file, but Audacity says 22.5 seconds). I've tried a variety of import format settings. What am I likely doing wrong here ?

  


  I'm a little new to working with audio, so I may be misunderstanding something.

  


  Edit : I tried panning the audio to the right channel, and its reflected in the data. It also looks like a repeating pattern exactly 1024 samples apart, which indicates to me a programming error with a buffer not getting overwritten after the first sample.


  


• How to pass through complex command line arguments in Process.Arguments ?

  5 juin 2015, par Louis de la Verité

  In C#, I have a complex set of arguments to pass through to Process.Arguments. I have circumvented the problem by using a one-line script of compiled Powershell. The exact "train" of arguments is :

  ffmpeg -framerate $fr -f image2 -i $temp\JPEG%01d -c:v libx264 -threads $threads -crf $encqual $video_out

  Problematic arguments are the -i (with the JPEG%01 pattern) & the -c argument. I have tried double escaping, but it’s not working any way I try. I don’t like the /<arg>: <val></val></arg> notation because it conflicts with the -c:v. As mentioned, I got it working by pass it off to PowerShell, but that’s a hack. The directory also gets parsed because it has spaces.

  ERROR :

  "Unable to find a suitable output format for ’C :\Program’
  "C :\Program" Invalid argument.

  CODE :

  Process proc = new Process();
  
  // proc.StartInfo = _fw.CreateProcessStartInfo4Ffmpeg(
  
        pathTempDir,    output);
  
  proc.StartInfo = _fw.CreateProcessStartInfo4AviGen(pathTempDir, output);
  
  proc.Start();
  
  
  
  // doesn't work yet
  
  public ProcessStartInfo CreateProcessStartInfo4Ffmpeg(string tempFile,
  
         string output)
  
  {
  
     string args = FormCmdArgsForFfmpeg(this, tempFile, output);
  
     ProcessStartInfo psi = CreateProcessStartInfo4Ffmpeg(this, args);
  
  return psi;
  
  }
  
  
  
  public static ProcessStartInfo CreateProcessStartInfo4Ffmpeg(
  
          FfmpegWrapper fw, string args)
  
   {
  
     string sExePath = clsRegistry.regInfo.FfmpegExePath;
  
     ProcessStartInfo pi = new ProcessStartInfo(sExePath);
  
  
  
     pi.UseShellExecute = false;
  
     pi.CreateNoWindow = false;
  
     pi.WindowStyle = ProcessWindowStyle.Normal;
  
  
  
     pi.Arguments = String.Copy(args);
  
     string sCmdFmtdMsg = string.Format("Command: ", 
  
             sExePath, pi.Arguments);
  
  return pi;
  
  } // end CreateProcessStartInfo4AviGen()
  
  
  
  // N.B. doesn't work yet; args prob
  
  public static string FormCmdArgsForFfmpeg(FfmpegWrapper fw, 
  
       string    tempDir, string output)
  
  {
  
   /* -framerate $framerate -f image2 -i $temp_dir_jpgs\JPEG%01d 
  
   -c:v libx264 -threads $threads -crf $encode_quality $video_out
  
   */
  
      string sFmtdArg2 = "-{0} \"{1}\" ";
  
      string sFmtdArg1 = "-{0} ";
  
      string sArgs = String.Empty;
  
      string sout = null;
  
      string pathInput = Path.Combine(tempDir, "JPEG%01d");
  
  
  
      StringBuilder sb = new StringBuilder("");
  
      sb.Append(string.Format(sFmtdArg2, "framerate", fw.framerate));
  
      sb.Append(string.Format(sFmtdArg2, "f", "image2"));
  
      sb.Append(string.Format(sFmtdArg2, "i", pathInput));
  
      sb.Append(string.Format(sFmtdArg2, "c:v", "libx264"));
  
      sb.Append(string.Format(sFmtdArg2, "threads", fw.threads) );
  
      sb.Append(string.Format(sFmtdArg2, "crf", fw.encode_quality));
  
      sb.Append(output);
  
      sout = sb.ToString();
  
  return sout;
  
  }